This study focuses on preventing suicide and suicide behaviors among vulnerable youth--identified alcohol and drug users and potential school dropouts. A disturbing trend among these youths is the simultaneous occurrence of substance use/abuse, depression, suicide ideation and attempts, and poor school performance often leading to high school dropout. They are similar to 25% of the nation's youth whose health is threatened by multiple risk-taking behavioral and social lifestyles. Thus, scholars call for empirically tested strategies for preventing suicide and other risky behaviors. Accordingly, we propose a special high school-based prevention program for vulnerable youth evidencing suicide- risk thoughts and behaviors. An integrated social network support/life skills training program is designed to address multiple risk factors such as prior environmental stress, social disorganization, skill deficiencies and high-risk behaviors. A year-long, class with a teacher/student ratio of 1:12 will provide program students with opportunities for life-skills training and social network development. Two specific aims are 1) to test, in a school-based field experiment, the effectiveness of the program in achieving four outcomes: less suicide ideation/behaviors, less depression, less drug involvement and less school deviance; and 2) to test a causal model for the roles of teacher, peer, and school network support/skills training in counteracting antecedent suicide-risk factors. We predict that the teacher and peer-group intervention will increase life skills competencies and prosocial network bonding and account for reductions in depression and drug involvement and, thereby, decrease suicide lethality and school deviance. Study subjects will be 126-172 randomly selected vulnerable youths at-risk for suicide behaviors (experimental and controls), and approximately 200 non-suicidal vulnerable youths enrolled in an existing NIDA funded project. Repeated measures ANCOVA, trend analysis, and a causal modeling approach to the experimental design will be used to evaluate the program in five high schools. Structural equation modeling with multiple indicators will be used to test competing models of the effects of personal and environmental risk factors and protective life skills competencies and social support resources on adolescents' suicidal behaviors, depression, drug involvement, and school performance over time. The study should increase our understanding of the risk and protective factors involved in the causes and prevention of adolescent suicidal behaviors; and of how prevention strategies work to decrease high-risk behavioral lifestyles among youth. Long range, the study should reveal particular prevention program elements that are linked to decreasing the likelihood of suicide: the third leading cause of death among America's youth.